POULTRY WORK. 207 



utility points could not be bred into standard varieties of poul- 

 try. 



TWO EXPERIMENTS PLANNED. 

 The general plan of the investigation on this matter was sim- 

 ply to try on a large scale selection experiments in egg produc- 

 tion. The working basis of the original plan was the assump- 

 tion that small fluctuating variations in egg production were 

 inherited. Proceeding on this assumption two experiments 

 were planned by the writer. One of these experiments had 

 as its object to determine whether by breeding only from rela- 

 tively high layers the average annual egg production of the flock 

 could be increased. This experiment was carried out by Pro- 

 fessor Gowell through 9 consecutive years. The limit of "rela- 

 tively high laying" was taken at 150 eggs. All birds laying more 

 than that in their first laying year were used as breeders during 

 the 9 years which this experiment was continued. The cock- 

 erels used in this breeding experiment were sons of birds laying 

 200 or more eggs in their first laying year. The second experi- 

 ment planned which was of obviously equal importance to the 

 first, was intended to determine whether if only relatively poor 

 laying hens were used as breeders it would be possible to lower 

 the annual average egg production. This second experiment 

 was never tried, a fact which is greatly to be regretted in the 

 light of what is now known regarding the results of the first 

 experiment (see p. 210 below). A word may perhaps be advis- 

 able to point out clearly why it was necessary to perform this 

 second experiment before any definite conclusions could be 

 drawn as to the meaning of any results obtained in the first 

 experiment. Suppose that in trying the first experiment favor- 

 able results are obtained, i. e., that the annual average produc- 

 tion increases from year to year concurrently with the selective 

 breeding. Is it to be concluded that the breeding is the cause 

 of the increase? Such a conclusion obviously cannot be drawn, 

 if, as was the case in the actual performance of this "breeding 

 for egg production" work at this Station, there are made 

 improvements in housing, feeding and other details of manage- 

 ment, each one of which by itself might tend to increase annual 

 average egg production. Nobody can tell how much of the 

 observed improvement is due to breeding and how much to 

 housing, feeding, etc. 



